Start
A wild obstacle appears
Cleanup
- Excel for initial exploration and some metadata
- Data cleaning using R
How to navigate?
Land Ahoy!
- Something interesting related to waiting time?
- Case severity?
- Regions?
- Sex!
No acute, only lateral + primary operation
#
What does this mean?
- If two groups of each 100 men and women have a primary lateral inguinal hernia
- After 100 days, 53 women and 46 men waiting, 7 more women
- After 150 days, 35 women and 26 men waiting, 9 more women
- After 200 days, 20 women and 15 men waiting, 5 more women
A lot to explore…
- Coxph model shows ~18% higher hazard rate for men (95% CI 13% - 28%, age adjusted)
- Proportional hazards ok in base case
- Explorative, not properly looked at biases/causal pathway
- -> Be careful to draw conclusions!
Conclusions to bring home?
- Does it really matter?
Thanks & Resources
- Course coordinater, administrator and all lecturers
- Henrik Holmberg
- R, RStudio, and contributors
Sail away
Copyright info:
https://commons.wikimedia.org/wiki/File:WIKIMEDIA_15ANS_SWITZ_HELVETIA-01-01.svg Stefan Schlageter
https://commons.wikimedia.org/wiki/File:R_logo.svg Hadley Wickham and others at RStudio
https://commons.wikimedia.org/wiki/File:Microsoft_Excel_Logo.svg Microsoft
https://github.com/jakeybob/plotly-survival/blob/master/examples.R PlotlySurvival
Results II
| N | Median | Mean | |
|---|---|---|---|
| Kvinna | 1012 | 106 | 129.2698 |
| Man | 13114 | 91 | 112.4308 |
Results III
| Dependent variable: | |
| vantetid | |
| SEXMan | 0.186*** |
| (0.033) | |
| age | -0.003*** |
| (0.001) | |
| Observations | 14,126 |
| R2 | 0.004 |
| Max. Possible R2 | 1.000 |
| Log Likelihood | -120,837.700 |
| Wald Test | 52.370*** (df = 2) |
| LR Test | 53.751*** (df = 2) |
| Score (Logrank) Test | 52.438*** (df = 2) |
| Note: | p<0.1; p<0.05; p<0.01 |
| stargazer html | |